import pandas as pd              # Used for data manipulation and analysis
import numpy as np               # Used for numerical operations and arrays
import matplotlib.pyplot as plt  # Import the matplotlib library's pyplot module for plotting data visualizations
import seaborn as sns            # Import the seaborn library for advanced data visualization
import ast                       # Import the ast module for abstract syntax tree operations

from sklearn.model_selection import train_test_split               # Used to split data into training and testing sets
from sklearn.svm import SVC                                        # SVC for classification
from sklearn.metrics import accuracy_score, classification_report  # Metrics to evaluate model performance
from sklearn.metrics import ConfusionMatrixDisplay                 # used to display confusion matrix
from sklearn.preprocessing import StandardScaler                   # Import the StandardScaler from scikit-learn's preprocessing module

df = pd.read_csv("week8_data.csv")
df.head(10)

print(df.shape)   #Print the shape

(2000, 4)

df.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2000 entries, 0 to 1999
Data columns (total 4 columns):
 #   Column    Non-Null Count  Dtype 
---  ------    --------------  ----- 
 0   x_values  2000 non-null   object
 1   y_values  2000 non-null   object
 2   z_values  2000 non-null   object
 3   label     2000 non-null   object
dtypes: object(4)
memory usage: 62.6+ KB

print(list(df.columns))    # Print a list of column names

['x_values', 'y_values', 'z_values', 'label']

missing_values = df.isnull().sum()   # Calculate the total number of missing values
print(missing_values)                  # Print the number of missing values for each column to identify data completeness issues

x_values    0
y_values    0
z_values    0
label       0
dtype: int64

label_counts = df['label'].value_counts()
label_counts

label
walking    1000
sitting    1000
Name: count, dtype: int64

plt.figure(figsize=(6, 4))
label_counts.plot(kind='bar')
plt.title('Distribution of Labels')
plt.xlabel('Label')
plt.ylabel('Count')
plt.show()

df['x_values'] = df['x_values'].apply(ast.literal_eval)  # Converts strings in 'x_values' column to lists
df['y_values'] = df['y_values'].apply(ast.literal_eval)  # Converts strings in 'y_values' column to lists
df['z_values'] = df['z_values'].apply(ast.literal_eval)  # Converts strings in 'z_values' column to lists

df

# Function to calculate the mean of a list
def calculate_mean(lst):
    return sum(lst) / len(lst)  # Returns the average value of the list

# Function to calculate the range of values in a list
def calculate_range(lst):
    return max(lst) - min(lst)  # Returns the difference between the max and min values

# Function to calculate the correlation between two lists
def calculate_correlation(list1, list2):
    return np.corrcoef(list1, list2)[0, 1]  # Returns the Pearson correlation coefficient between two lists

df['x_mean'] = df['x_values'].apply(calculate_mean)  # Calculates the mean of x-axis values
df['y_mean'] = df['y_values'].apply(calculate_mean)  # Calculates the mean of y-axis values
df['z_mean'] = df['z_values'].apply(calculate_mean)  # Calculates the mean of z-axis values

df

df['x_range'] = df['x_values'].apply(calculate_range)  # Calculates the range of x-axis values
df['y_range'] = df['y_values'].apply(calculate_range)  # Calculates the range of y-axis values
df['z_range'] = df['z_values'].apply(calculate_range)  # Calculates the range of z-axis values

df

df['xy_corr'] = df.apply(lambda row: calculate_correlation(row['x_values'], row['y_values']), axis=1)  # Correlation between x and y values
df['yz_corr'] = df.apply(lambda row: calculate_correlation(row['y_values'], row['z_values']), axis=1)  # Correlation between y and z values
df['xz_corr'] = df.apply(lambda row: calculate_correlation(row['x_values'], row['z_values']), axis=1)  # Correlation between x and z values

df

def encode_label(lbl):
    # Mapping dictionary to assign integer values to each label
    label_map = {'walking': 1, 'sitting': 0}
    # Return the encoded value for the label; if label not found, return -1
    return label_map.get(lbl, -1)

# Apply the encoding function to the 'label' column in the DataFrame
df['label'] = df['label'].apply(encode_label)

df

df = df.iloc[:, 3:]
df

# Plotting each column in separate subplots for a comprehensive overview
columns_to_plot = df.columns[1:]  # Excluding 'label' column
num_columns = len(columns_to_plot)

# Creating subplots for each feature in the DataFrame
fig, axes = plt.subplots(nrows=num_columns, ncols=1, figsize=(10, 2 * num_columns))
fig.suptitle("Distribution of Each Feature in the DataFrame", y=1.02)

# Plot each column in a separate subplot
for i, column in enumerate(columns_to_plot):
    axes[i].plot(df.index, df[column])
    axes[i].set_title(f"{column} Over Observations")
    axes[i].set_xlabel("Index")
    axes[i].set_ylabel(column)

plt.tight_layout()
plt.show()

# Calculate the correlation matrix for the DataFrame
correlation_matrix = df.corr()

# Plot a heatmap to visualize the correlation matrix
plt.figure(figsize=(10, 8))
sns.heatmap(correlation_matrix, annot=True, cmap="coolwarm", center=0, linewidths=0.5)
plt.title("Correlation Heatmap of Features")
plt.show()

# Plot pairwise relationships in the dataset
sns.pairplot(df, hue='label', diag_kind='kde', plot_kws={'alpha':0.5})
plt.suptitle("Pair Plot of Features Colored by Label", y=1.02)
plt.show()

X = df.iloc[:, 1:]   # All columns after the first one are features
y = df['label']      # The first column is the target (label)

X

y

0       1
1       1
2       1
3       1
4       1
       ..
1995    0
1996    0
1997    0
1998    0
1999    0
Name: label, Length: 2000, dtype: int64

scaler = StandardScaler()               # Initialize the StandardScaler
X_normalized = scaler.fit_transform(X) # Fit the scaler to the features (X) and transform them to standardize
X_normalized

array([[-0.77763893, -0.63530781,  0.90317761, ..., -0.2687097 ,
        -1.09678083, -0.2408278 ],
       [-0.76640983, -0.68909079,  0.91206351, ..., -0.60724429,
        -0.68577841, -0.56632177],
       [-0.76517924, -0.67819803,  0.91276874, ...,  0.41121031,
        -0.73784045, -1.01132414],
       ...,
       [ 0.95779658,  0.79623857, -0.2199722 , ..., -0.36658596,
        -0.32438925, -0.69805523],
       [ 0.9394916 ,  0.81513069, -0.22603718, ..., -1.09105009,
        -0.06638503, -0.02489821],
       [ 0.88365371,  0.90346413, -0.29627812, ..., -1.77642952,
        -1.38540358,  0.97810711]])

X_train, X_test, y_train, y_test = train_test_split(X_normalized, y, test_size=0.2)  # Using 80% of the samples for training, and 20% for testing

print("X_train shape is : ", X_train.shape)
print("X_test shape  is : ", X_test.shape)
print("y_train shape is : ", y_train.shape)
print("y_test shape is  : ", y_test.shape)

X_train shape is :  (1600, 9)
X_test shape  is :  (400, 9)
y_train shape is :  (1600,)
y_test shape is  :  (400,)

svm_model = SVC(kernel='poly')

svm_model.fit(X_train, y_train)

SVC(kernel='poly')

SVC(kernel='poly')

y_pred = svm_model.predict(X_test)

print("Accuracy:", accuracy_score(y_test, y_pred))
print("Classification Report:\n", classification_report(y_test, y_pred))

Accuracy: 0.9125
Classification Report:
               precision    recall  f1-score   support

           0       0.88      0.96      0.92       215
           1       0.95      0.85      0.90       185

    accuracy                           0.91       400
   macro avg       0.92      0.91      0.91       400
weighted avg       0.92      0.91      0.91       400

# Display a confusion matrix to visualize the classification performance
fig, ax = plt.subplots(figsize=(8, 6))
ConfusionMatrixDisplay.from_predictions(y_test, y_pred, ax=ax, cmap="Blues", colorbar=False)
plt.title("Confusion Matrix")

Text(0.5, 1.0, 'Confusion Matrix')

import pickle   # Import the pickle module for saving and loading Python objects

# the file is essentially a stream of bytes; it's common to use .pkl as the file extension
with open('svm_model.pkl', 'wb') as file:
    pickle.dump(svm_model, file)

# we will also need the same scaler which was used to train the model
with open('scaler.pkl', 'wb') as file:
    pickle.dump(scaler, file)

# Run this cell to check the version of scikit-learn on your PC

import sklearn
sklearn.__version__

'1.5.1'

# A dummy data point created with arbitary values

import warnings
import pickle
import numpy as np

warnings.filterwarnings("ignore")

with open('svm_model.pkl', 'rb') as file:
    model = pickle.load(file)

with open('scaler.pkl', 'rb') as file:
    scaler = pickle.load(file)
    
features = np.array([6.0, 6.3, 2.1, 2.5, 1.3, 4.6, 0, 0, -0.1,]).reshape(1, -1)

# use the same scaler to transform data, as we did in training
features_normalized = scaler.transform(features)

# make a single prediction using fake features
predicted_label = model.predict(features_normalized)[0]

if predicted_label == 0:
    print("Walking")
elif predicted_label == 1:
    print("Sitting")
else:
    print("Unknown activity")

Sitting

import warnings
import pickle
import numpy as np
from pymongo import MongoClient
import time

warnings.filterwarnings("ignore")

# Load the model and scaler
with open('svm_model.pkl', 'rb') as file:
    model = pickle.load(file)

with open('scaler.pkl', 'rb') as file:
    scaler = pickle.load(file)

# Connect to MongoDB
client = MongoClient("mongodb+srv://stanzir...........") # Replace MongoDB URI connection string
db = client["HAR"]  # Replace Database name
collection = db["Lab_Web"] # Replace collection name

# Global variable to store the latest prediction
latest_prediction = "Waiting for data..."

def calculate_features(data):
    x_values = np.array(data.get('x_values', []))
    y_values = np.array(data.get('y_values', []))
    z_values = np.array(data.get('z_values', []))

    if x_values.size > 0 and y_values.size > 0 and z_values.size > 0:
        x_mean = np.mean(x_values)
        y_mean = np.mean(y_values)
        z_mean = np.mean(z_values)
        x_range = np.max(x_values) - np.min(x_values)
        y_range = np.max(y_values) - np.min(y_values)
        z_range = np.max(z_values) - np.min(z_values)
        xy_corr = np.corrcoef(x_values, y_values)[0, 1] if x_values.size > 1 else 0
        yz_corr = np.corrcoef(y_values, z_values)[0, 1] if y_values.size > 1 else 0
        xz_corr = np.corrcoef(x_values, z_values)[0, 1] if z_values.size > 1 else 0
        return np.array([[x_mean, y_mean, z_mean, x_range, y_range, z_range, xy_corr, yz_corr, xz_corr]])
    else:
        return np.zeros((1, 9))

def fetch_and_predict():
    global latest_prediction
    last_id = None

    try:
        while True:
            query = {} if last_id is None else {"_id": {"$gt": last_id}}
            data = collection.find_one(query, sort=[("_id", 1)])

            if data:
                last_id = data["_id"]
                features = calculate_features(data)
                features_normalized = scaler.transform(features)
                predicted_label = model.predict(features_normalized)[0]

                if predicted_label == 0:
                    latest_prediction = "Walking"
                elif predicted_label == 1:
                    latest_prediction = "Sitting"
                else:
                    latest_prediction = "Unknown activity"

                print(f"Prediction for document {last_id}: {latest_prediction}")
            else:
                print("No new data found. Waiting...")
                time.sleep(5)

    except KeyboardInterrupt:
        print("Prediction loop interrupted. Exiting gracefully.")

# Start the prediction loop
fetch_and_predict()

Prediction for document 67249958c328992061f010da: Sitting
Prediction for document 67249958c328992061f010db: Walking
Prediction for document 67249958c328992061f010dc: Sitting
Prediction for document 67249958c328992061f010dd: Walking
Prediction for document 67249958c328992061f010de: Sitting
Prediction for document 67249958c328992061f010df: Sitting
Prediction for document 67249958c328992061f010e0: Walking
Prediction for document 67249958c328992061f010e1: Sitting
Prediction for document 67249958c328992061f010e2: Sitting
Prediction for document 67249958c328992061f010e3: Walking
Prediction for document 67249958c328992061f010e4: Sitting
Prediction for document 67249958c328992061f010e5: Sitting
Prediction for document 67249958c328992061f010e6: Walking
Prediction for document 67249958c328992061f010e7: Sitting
Prediction for document 67249958c328992061f010e8: Sitting
Prediction for document 67249958c328992061f010e9: Walking
Prediction for document 67249958c328992061f010ea: Walking
Prediction for document 67249958c328992061f010eb: Walking
Prediction for document 67249958c328992061f010ec: Walking
No new data found. Waiting...
Prediction loop interrupted. Exiting gracefully.

	x_values	y_values	z_values	label
0	[-1.2552512, -1.2160246, -1.0983448, -1.216024...	[1.6475172, 1.2160246, 1.6475172, 0.9414384, 1...	[8.7867584, 8.7867584, 8.747531799999999, 8.82...	walking
1	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2944778, 1.1767979999999998, 1.3337044, 1.1...	[8.8652116, 8.8652116, 8.825985, 8.9044382, 8....	walking
2	[-1.0983448, -1.0983448, -1.1375714, -1.098344...	[1.0591182, 1.2160246, 1.2944778, 1.0983448, 1...	[8.9044382, 8.8652116, 8.8652116, 8.8652116, 8...	walking
3	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2552512, 1.1375714, 1.1375714, 1.1767979999...	[8.7867584, 8.825985, 8.825985, 8.825985, 8.82...	walking
4	[0.2353596, 0.2353596, 0.3138128, 0.2745862, 0...	[-0.196133, -0.196133, -0.196133, -0.2353596, ...	[9.1005712, 9.0613446, 9.0613446, 9.1005712, 9...	walking
5	[0.2353596, 0.2745862, 0.2353596, 0.3138128, 0...	[-0.196133, -0.2353596, -0.1569064, -0.196133,...	[9.1005712, 9.1397978, 9.0613446, 9.0613446, 9...	walking
6	[0.2353596, 0.3138128, 0.196133, 0.2745862, 0....	[-0.1569064, -0.2353596, -0.1569064, -0.235359...	[9.022118, 8.9828914, 9.022118, 9.1397978, 9.0...	walking
7	[3.8834334, 4.0795664, 4.158019599999999, 3.33...	[7.4138274, 4.314926, 5.4917240000000005, -1.5...	[8.9436648, 7.8845466, 9.4536106, 10.002782999...	walking
8	[0.3530394, 0.392266, 0.4707192, 0.392266, 0.3...	[-0.1569064, -0.0784532, -0.1176798, -0.078453...	[9.0613446, 8.9828914, 9.1005712, 9.1397978, 9...	walking
9	[1.2552512, 1.1375714, 0.3138128, -0.1569064, ...	[9.4536106, 9.4536106, 9.3751574, 9.2574776, 8...	[2.1182364, 2.157463, 1.8828768, 1.5298374, 1....	walking

	x_values	y_values	z_values	label
0	[-1.2552512, -1.2160246, -1.0983448, -1.216024...	[1.6475172, 1.2160246, 1.6475172, 0.9414384, 1...	[8.7867584, 8.7867584, 8.747531799999999, 8.82...	walking
1	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2944778, 1.1767979999999998, 1.3337044, 1.1...	[8.8652116, 8.8652116, 8.825985, 8.9044382, 8....	walking
2	[-1.0983448, -1.0983448, -1.1375714, -1.098344...	[1.0591182, 1.2160246, 1.2944778, 1.0983448, 1...	[8.9044382, 8.8652116, 8.8652116, 8.8652116, 8...	walking
3	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2552512, 1.1375714, 1.1375714, 1.1767979999...	[8.7867584, 8.825985, 8.825985, 8.825985, 8.82...	walking
4	[0.2353596, 0.2353596, 0.3138128, 0.2745862, 0...	[-0.196133, -0.196133, -0.196133, -0.2353596, ...	[9.1005712, 9.0613446, 9.0613446, 9.1005712, 9...	walking
...	...	...	...	...
1995	[6.2370294, 6.1978028, 6.1585762, 6.1978028, 6...	[6.7469752000000005, 6.864655, 6.7862018000000...	[3.4911673999999997, 3.5696205999999995, 3.569...	sitting
1996	[6.2370294, 6.276256, 6.1585762, 6.3154826, 6....	[6.668522, 6.8254284, 7.1000146, 6.5116156, 6....	[3.6088471999999996, 3.6480737999999997, 3.530...	sitting
1997	[6.1978028, 6.1585762, 6.1585762, 6.2370294, 6...	[6.7077486, 6.864655, 6.786201800000001, 6.903...	[3.6088471999999996, 3.6480737999999997, 3.648...	sitting
1998	[6.1978028, 6.2370294, 6.1978028, 6.1585762, 6...	[6.864655, 6.786201800000001, 6.78620180000000...	[3.6088471999999996, 3.5696205999999995, 3.530...	sitting
1999	[6.3154826, 6.1585762, 6.2370294, 6.276256, 6....	[6.982334799999999, 6.668522, 6.6292954, 6.590...	[3.4911673999999997, 3.5696205999999995, 3.530...	sitting

	x_values	y_values	z_values	label	x_mean	y_mean	z_mean
0	[-1.2552512, -1.2160246, -1.0983448, -1.216024...	[1.6475172, 1.2160246, 1.6475172, 0.9414384, 1...	[8.7867584, 8.7867584, 8.747531799999999, 8.82...	walking	-1.196411	1.313437	8.807025
1	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2944778, 1.1767979999999998, 1.3337044, 1.1...	[8.8652116, 8.8652116, 8.825985, 8.9044382, 8....	walking	-1.148686	1.106844	8.848213
2	[-1.0983448, -1.0983448, -1.1375714, -1.098344...	[1.0591182, 1.2160246, 1.2944778, 1.0983448, 1...	[8.9044382, 8.8652116, 8.8652116, 8.8652116, 8...	walking	-1.143455	1.148686	8.851482
3	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2552512, 1.1375714, 1.1375714, 1.1767979999...	[8.7867584, 8.825985, 8.825985, 8.825985, 8.82...	walking	-1.117304	1.165684	8.823370
4	[0.2353596, 0.2353596, 0.3138128, 0.2745862, 0...	[-0.196133, -0.196133, -0.196133, -0.2353596, ...	[9.1005712, 9.0613446, 9.0613446, 9.1005712, 9...	walking	0.271971	-0.209209	9.054153
...	...	...	...	...	...	...	...
1995	[6.2370294, 6.1978028, 6.1585762, 6.1978028, 6...	[6.7469752000000005, 6.864655, 6.7862018000000...	[3.4911673999999997, 3.5696205999999995, 3.569...	sitting	6.158576	6.832620	3.602309
1996	[6.2370294, 6.276256, 6.1585762, 6.3154826, 6....	[6.668522, 6.8254284, 7.1000146, 6.5116156, 6....	[3.6088471999999996, 3.6480737999999997, 3.530...	sitting	6.191265	6.794047	3.607540
1997	[6.1978028, 6.1585762, 6.1585762, 6.2370294, 6...	[6.7077486, 6.864655, 6.786201800000001, 6.903...	[3.6088471999999996, 3.6480737999999997, 3.648...	sitting	6.179497	6.812353	3.601002
1998	[6.1978028, 6.2370294, 6.1978028, 6.1585762, 6...	[6.864655, 6.786201800000001, 6.78620180000000...	[3.6088471999999996, 3.5696205999999995, 3.530...	sitting	6.101698	6.884922	3.572889
1999	[6.3154826, 6.1585762, 6.2370294, 6.276256, 6....	[6.982334799999999, 6.668522, 6.6292954, 6.590...	[3.4911673999999997, 3.5696205999999995, 3.530...	sitting	5.864377	7.224232	3.247309

	x_values	y_values	z_values	label	x_mean	y_mean	z_mean	x_range	y_range	z_range
0	[-1.2552512, -1.2160246, -1.0983448, -1.216024...	[1.6475172, 1.2160246, 1.6475172, 0.9414384, 1...	[8.7867584, 8.7867584, 8.747531799999999, 8.82...	walking	-1.196411	1.313437	8.807025	0.274586	1.216025	0.353039
1	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2944778, 1.1767979999999998, 1.3337044, 1.1...	[8.8652116, 8.8652116, 8.825985, 8.9044382, 8....	walking	-1.148686	1.106844	8.848213	0.274586	1.059118	0.235360
2	[-1.0983448, -1.0983448, -1.1375714, -1.098344...	[1.0591182, 1.2160246, 1.2944778, 1.0983448, 1...	[8.9044382, 8.8652116, 8.8652116, 8.8652116, 8...	walking	-1.143455	1.148686	8.851482	0.431493	0.784532	0.431493
3	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2552512, 1.1375714, 1.1375714, 1.1767979999...	[8.7867584, 8.825985, 8.825985, 8.825985, 8.82...	walking	-1.117304	1.165684	8.823370	0.235360	0.902212	0.313813
4	[0.2353596, 0.2353596, 0.3138128, 0.2745862, 0...	[-0.196133, -0.196133, -0.196133, -0.2353596, ...	[9.1005712, 9.0613446, 9.0613446, 9.1005712, 9...	walking	0.271971	-0.209209	9.054153	0.156906	0.235360	0.235360
...	...	...	...	...	...	...	...	...	...	...
1995	[6.2370294, 6.1978028, 6.1585762, 6.1978028, 6...	[6.7469752000000005, 6.864655, 6.7862018000000...	[3.4911673999999997, 3.5696205999999995, 3.569...	sitting	6.158576	6.832620	3.602309	0.627626	1.098345	1.019892
1996	[6.2370294, 6.276256, 6.1585762, 6.3154826, 6....	[6.668522, 6.8254284, 7.1000146, 6.5116156, 6....	[3.6088471999999996, 3.6480737999999997, 3.530...	sitting	6.191265	6.794047	3.607540	0.313813	0.588399	0.353039
1997	[6.1978028, 6.1585762, 6.1585762, 6.2370294, 6...	[6.7077486, 6.864655, 6.786201800000001, 6.903...	[3.6088471999999996, 3.6480737999999997, 3.648...	sitting	6.179497	6.812353	3.601002	0.235360	0.313813	0.431493
1998	[6.1978028, 6.2370294, 6.1978028, 6.1585762, 6...	[6.864655, 6.786201800000001, 6.78620180000000...	[3.6088471999999996, 3.5696205999999995, 3.530...	sitting	6.101698	6.884922	3.572889	0.470719	0.980665	0.392266
1999	[6.3154826, 6.1585762, 6.2370294, 6.276256, 6....	[6.982334799999999, 6.668522, 6.6292954, 6.590...	[3.4911673999999997, 3.5696205999999995, 3.530...	sitting	5.864377	7.224232	3.247309	1.137571	2.392823	1.608291

	x_values	y_values	z_values	label	x_mean	y_mean	z_mean	x_range	y_range	z_range	xy_corr	yz_corr	xz_corr
0	[-1.2552512, -1.2160246, -1.0983448, -1.216024...	[1.6475172, 1.2160246, 1.6475172, 0.9414384, 1...	[8.7867584, 8.7867584, 8.747531799999999, 8.82...	walking	-1.196411	1.313437	8.807025	0.274586	1.216025	0.353039	-0.098106	-0.526352	-0.182488
1	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2944778, 1.1767979999999998, 1.3337044, 1.1...	[8.8652116, 8.8652116, 8.825985, 8.9044382, 8....	walking	-1.148686	1.106844	8.848213	0.274586	1.059118	0.235360	-0.240468	-0.340499	-0.322237
2	[-1.0983448, -1.0983448, -1.1375714, -1.098344...	[1.0591182, 1.2160246, 1.2944778, 1.0983448, 1...	[8.9044382, 8.8652116, 8.8652116, 8.8652116, 8...	walking	-1.143455	1.148686	8.851482	0.431493	0.784532	0.431493	0.187816	-0.364041	-0.513297
3	[-1.1767979999999998, -1.1767979999999998, -1....	[1.2552512, 1.1375714, 1.1375714, 1.1767979999...	[8.7867584, 8.825985, 8.825985, 8.825985, 8.82...	walking	-1.117304	1.165684	8.823370	0.235360	0.902212	0.313813	0.091160	-0.215989	-0.238869
4	[0.2353596, 0.2353596, 0.3138128, 0.2745862, 0...	[-0.196133, -0.196133, -0.196133, -0.2353596, ...	[9.1005712, 9.0613446, 9.0613446, 9.1005712, 9...	walking	0.271971	-0.209209	9.054153	0.156906	0.235360	0.235360	0.109223	-0.355972	-0.468424
...	...	...	...	...	...	...	...	...	...	...	...	...	...
1995	[6.2370294, 6.1978028, 6.1585762, 6.1978028, 6...	[6.7469752000000005, 6.864655, 6.7862018000000...	[3.4911673999999997, 3.5696205999999995, 3.569...	sitting	6.158576	6.832620	3.602309	0.627626	1.098345	1.019892	-0.222889	-0.062302	-0.477676
1996	[6.2370294, 6.276256, 6.1585762, 6.3154826, 6....	[6.668522, 6.8254284, 7.1000146, 6.5116156, 6....	[3.6088471999999996, 3.6480737999999997, 3.530...	sitting	6.191265	6.794047	3.607540	0.313813	0.588399	0.353039	-0.482589	-0.040348	-0.378993
1997	[6.1978028, 6.1585762, 6.1585762, 6.2370294, 6...	[6.7077486, 6.864655, 6.786201800000001, 6.903...	[3.6088471999999996, 3.6480737999999997, 3.648...	sitting	6.179497	6.812353	3.601002	0.235360	0.313813	0.431493	-0.139266	-0.177081	-0.378796
1998	[6.1978028, 6.2370294, 6.1978028, 6.1585762, 6...	[6.864655, 6.786201800000001, 6.78620180000000...	[3.6088471999999996, 3.5696205999999995, 3.530...	sitting	6.101698	6.884922	3.572889	0.470719	0.980665	0.392266	-0.443920	-0.060412	-0.089779
1999	[6.3154826, 6.1585762, 6.2370294, 6.276256, 6....	[6.982334799999999, 6.668522, 6.6292954, 6.590...	[3.4911673999999997, 3.5696205999999995, 3.530...	sitting	5.864377	7.224232	3.247309	1.137571	2.392823	1.608291	-0.732138	-0.656865	0.340857

Week 8: Deploying Machine-Learning Solutions in IoT

Import necessary libraries¶

Import specific modules from scikit-learn for model creation and evaluation¶

Data Overview¶

Load and display the dataset into a pandas DataFrame¶

Print the shape of the dataset to understand its dimensions (rows, columns)¶

Display a concise summary of the DataFrame¶

Print a list of column names in the dataset to understand the available features¶

Calculate the total number of missing values in each column of the dataset¶

Count the occurrences of each unique value in the 'label' column¶

Plotting the label counts as a bar chart¶

Convert string representations of lists back into actual Python lists¶

Feature Extraction¶

Calculate the mean of acceleration values in each axis and add them as new columns¶

Calculate the range of acceleration values in each axis and add them as new columns¶

Calculate the correlations between the x, y, and z axes and add them as new columns¶

Define a function to encode labels based on predefined categories¶

Select all columns starting from the 4th column onward¶

Trends and patterns across all values¶

Generating a sample correlation matrix¶

Pair Plot¶

Machine Learning¶

Separate the features and target variable from the dataset¶

Normalize the data¶

Split the dataset into training and testing sets¶

Check the shape of all of these¶

ML Model¶

Create an SVM classifier with a polynomial kernel¶

Train the SVM model¶

Evaluate the model performance¶

Display a confusion matrix to visualize the classification performance¶

Save the trained model to a file¶

Run this on your Pi¶

To use the deployed model and make a single prediction¶

Continuously retrieve data from MongoDB and make predictions¶

Upload the mongodb_test_data.csv on MongoDB¶

Run this code on your Pi¶

Final Challenge: Real-Time Activity Prediction